Study On Bionic Moving Control And Dexterous Operating Planning For The Robot Astronaut In Space Station

With the rapid development of aeronautics and astronautics technology,China already has the ability to build large space station.However,in the hostile space environment,pressure from both physical and psychological as well as the heavy tasks disturb the spaceman.Therefore,through analyzing the kinetic mechanism and operate patterns of the spacemen,the method that utilizing the robot spaceman to replace the real spaceman is proposed,which can reduce cost and improve security.This dissertation prevents the stability control,bionic motion and dexterous operations for the robot spaceman in the space station with microgravity.Also the biological image processing under the nonlinear illumination conditions is proposed.This dissertation is supported by the project from National Natural Science Foundation of China and 921 manned space project.The main research issues are summarized as follow.Firstly,the complex conditions in the space stations,including the microgravity environment,the small and complex structures make the robot more sensitive to disturbance from the inertia force,joint friction and contact force with surroundings.To address this issue,this dissertation first presents a multi contact point dynamics model under microgravity environment,and then proposes a double feedforward control method which combine the joint motor dynamics and multi contact dynamics.This method achieve the continuation of joint position,velocity and acceleration,speed up the response time as well as improve the stability of robot spaceman during the rapidly changes of joint torque.Secondly,for the requirement of robot astronaut wide operation in space station,this section puts forward the motor pattern which imitates the astronauts moving by handrails in body coordination.When the robot holds the handrails by two hands,the arm and the space station form a rigid closed chain.If there is even a small error in the closed chain,huge conflict forces will generate.The conflict may accumulate in the impedance controller and that will lead to robot instability.This article proposed an adaptive impedance control algorithm which changes the equilibrium position of virtual spring in impedance control timely.This algorithm is able to eliminate the conflict of internal force and achieve robot stably moving in a microgravity environment.Thirdly,when the robot astronaut performs the independent operation task,the apace complex light environment brings a huge challenge for robot vision system.To solve this problem,an improved algorithm of Retinex is put forward by Achimedean spiral to estimate the light intensity.This method improves the target recognition rate and location accuracy.There is another problem that the planning of operating types and sizes of tools is hard for multi-fingered hand.To solve this problem,a type of back-propagation neural network is built based on tools' shapes and sizes.Lastly,based on the prototype robot astronaut,a combined simulation system formed by Matlab and Adams is built,as well as the robot experiment platform.The experiment in the simulation system verifies the feasibility of the stability control algorithm under microgravity environment.The robot experiment platform verifies the effectiveness of the proposed research contents and methods for robot astronaut bionic moving and flexible operation.